What's in the dataset
Employment class action settlements approved by Washington, Oregon, and California courts: currently King, Snohomish, and Spokane County Superior Courts, both Washington federal districts (W.D. Wash. and E.D. Wash.), the District of Oregon, and the Central and Northern Districts of California (C.D. Cal. and N.D. Cal.), with additional counties and districts in progress. Coverage runs from 2017 to the present. "Employment" means wage-and-hour and pay-transparency/equal-pay class claims; consumer, privacy, and other non-employment class actions are excluded even when the class members happen to be employees.
Sources
Court records: motions for preliminary and final approval, approval orders, and fee orders, collected from court files and PACER/RECAP. Every financial figure is traced to a specific document; nothing is estimated or extrapolated.
Definitions
- Gross fund: the settlement fund as negotiated. For a claims-made settlement that is the agreed maximum, which can exceed what is finally paid out when claims come in under the cap. Every figure below is measured against that negotiated number, so a claims-made deal and a common fund of the same size are compared on the same basis.
- Fee %: the fee award actually approved by the court, as a percentage of the gross settlement fund (not the requested fee, and not net of costs).
- Days to final: calendar days from entry of the preliminary approval order to entry of the final approval order.
- Modified: the court made a substantive change to the requested terms (fees, service awards, notice plan, allocation). Ministerial edits don't count.
- Service award: the court-approved payment per named plaintiff / class representative.
- Claim type: the legal theories actually pleaded, read from the complaint and approval papers (meal & rest breaks, overtime, EPOA job-posting claims, RCW 49.62 moonlighting/noncompete classes, and so on). A settlement resolving several theories appears under each one, so claim-type counts intentionally sum past the settlement count. Statutes cited only as the vehicle for another theory don't count as their own claim type.
- Opt-in collectives: settlements of pure FLSA 216(b) collectives, where only workers who affirmatively opted in are bound and no Rule 23/CR 23 class was certified. These price differently from opt-out class settlements — the fund covers a self-selected group, not every absent class member — so the benchmarks exclude them by default. The "Opt-in collectives" filter adds them back, or shows them alone. Hybrid settlements (a Rule 23 class with a collective alongside) count as class settlements.
- Class size: the number of class members stated in the approval papers.
- Gross fund per class member: the gross settlement fund divided by class size. This is before fees, costs, and service awards come out; the checks class members actually receive are smaller.
- Approval year: the year the court entered the final approval order. Where a settlement has preliminary approval but the final order is not yet on file, the preliminary approval year is used so the deal counts once it exists. The current year is marked YTD because it is still filling in. Approval year rather than filing year: this is a dataset of approved settlements, so recent filing years would look artificially quiet only because those cases have not resolved yet.
- New filings: counts of new cases, not settlements, from the two courts where this dataset has complete filing coverage. King County Superior Court: cases in which plaintiff's counsel is one of the plaintiff-side employment class-action firms this dataset follows, across the civil case types those firms use. The court's own case-type codes do not identify class actions, and its Employment case type only dates from late 2023, so counsel is the one signal that reads the same across years. Western District of Washington: cases with an employment nature of suit, meaning 442 civil rights (jobs), 710 Fair Labor Standards Act, and 790 other labor, including cases removed from state court. Both series count class and individual actions, and a case removed from King County to the Western District appears in both. "Flagged as class actions" counts federal cases whose docket text identifies a class action so far; that identification lags filing by weeks to months, so the newest months read low. Only complete months are shown, and the card is updated monthly. Federal docket details post late, so a Western District month appears about six weeks after it ends and that court runs a month behind King County; earlier months can still be revised slightly.
- Quartiles: the first quartile, median, and third quartile of the gross fund among the settlements approved in a given year, computed by linear interpolation (the same method as a spreadsheet's QUARTILE function). Half of a year's settlements fall between the first and third quartile.
Why judges are anonymized here
The judge-level rows use consistent letters rather than names. The underlying orders are public record, but a ranked public table invites lazy readings of small samples. The named analysis, with the case-level context that makes it fair, is available in conversation. Get in touch.
Caveats
- Medians throughout. Subset sizes (n) are shown wherever a figure could mislead without one.
- Some rows await final orders; their financial fields are excluded until approval is entered.
- This is a growing dataset under active verification. Figures may shift from time to time.
About
Compiled and maintained by Rob Gillette, mediator and employment attorney, Gillette Mediation. Questions, corrections, or a case I've missed: contact me.